Artificial intelligence is a bustling concept that is emerging in the last few decades in ways we could not have imagined before. The face of modern AI, is diverse, machine centric, with wildly complex intelligent systems, promising to change every aspect of human life. These applications are found mostly in robot innovations, fueling a fascinating future for the field of robotics.
Demystifying the term
Although increasingly becoming part of our lives, John McCarthy’s term “Artificial Intelligence” in 1956 still sounds like a mythical future prediction more than a reality for the most of us. Perhaps this relates to the difficulty in defining the term. Artificial Intelligence is a theory that its ultimate form would entail the recreation of human thought process, operating through a man-made machine that shares our intellectual abilities. These would include features such as the ability to learn pretty much everything, the ability for reasoning, using language and formulating original ideas.
Artificial Intelligence and robotics: a long lasting relationship
Artificial intelligence has diachronically being linked with robotics as the two are closely related areas of study that fall under the field of computer science. This strong synergy between robotics and AI resulted in a rich research agenda on robotics platforms at their early beginning, which however progressed widely apart in the next decades (Sharma and Goswami, 2014). Nevertheless, with the proliferation of technologies, more mature techniques from both areas, more accessible robot platforms sharing advanced sensory motor capabilities, and a better understanding of the scientific challenges of the AI–Robotics intersection, we are now witnessing a revival in the interest and research in the fertile domain of embodied machine intelligence (Ingrand and Ghallab, 2012).
How does it work?
In understanding the relationship between AI and robots, one should think of AI as the software and robots manufactured as the hardware (REF). The connection between those two is that the control of the robot functions as a software agent sinceit reads data from the sensors, decides what to do next and then directs the effectors to act in the physical world (ref). AI is the brain, and the robot is its body, when a body does exist since there are cases like Siri where we hear only the woman’s voice, a personification of that AI, and there’s no robot involved at all.
When discussing AI and robotics interactions, in simple terms we refer to designing computers and machines that are acting similar to humans with minimum human involvement. The previous raises the challenge of unveiling the ways in which natural intelligence works. Developing AI is not a simple task of using a concrete model to work from. Neuroscience reveals how the brains consists of billions of neurons and our thoughts and learning stems from the electrical connections between different neurons. Yet we are still lacking the precise knowledge of how all of these connections make higher and even low logic operations. Based on this fact, AI research is largely theoretical. However the AI–Robotics intersection is very rich. According to Ingrand and Ghallab (2012:1), it covers issues such as:
• deliberate action, planning, acting, monitoring and goal reasoning,
• perceiving, modeling and understanding open environments,
• interacting with human and other robots,
• learning models required by the above functions, and
• integrating these functions in an adaptable and resilient architecture
What’s new in the field?
The rapid progress in artificial intelligence in robotics has involved advances such as deep learning, speech understanding and natural language processing, and artificial neural networks that are able to learn complex perceptual functions, likemathematical abstractions called deep convolutional neural networks, or ConvNets with capabilities to identify individuals from their faces, classify objects into one of 1,000 distinct categories etc.
Stemming from progress in these areas, the excitement in AI in robotics nowadays swirls aroundhow robots are now able apart from solving problems, learn to write, dream, play music, lie, read our emotions, and even deny carrying out orders with reasonable arguments. Below are only a few examples of research in the field that is taking place worldwide.
Of the most promising projects that scientists are now working include Microsoft’s Emotion API project, a cloud-based machine learning system designed to let A.I. read human emotions by analyzing our facial expressions. Equally important is the work of Gordon Briggs and Matthias Scheutz, from Tufts University’sHuman-Robot Interaction Lab, who are working to developmechanisms for robots to reject orders that it receives from humans, by effectively communicate why it rejected an order.
Meanwhile, researchers at UC Berkley have been using deep learning to train robots like BRETT (the Berkeley Robot for the Elimination of Tedious Tasks) to do every day human motor tasks that require dexterity and spatial awareness using algorithms, marking a major milestone in the field of artificial intelligence”. Further to these, at the University of California, Berkeley led by post-doctorate Igor Mordatch,have programmed a humanoid robot called Darwin, using a neural network to learn how to move and stand just like a child would learn.
Another breakthrough project involves a Boston company called Rethink Robotics which has already developed a pair of industrial robots that can work amongst and be trained by humans in a factory. Elsewhere in the US, MIT’s Computer Science and Artificial Intelligence Laboratory will receive “R5,” a six-foot, 290-pound humanoid robot also known as “Valkyrie” to be used for future space missions to Mars and beyond.
Despite fears that A.I will open the path for robots to become like humans and even surpass them, scientists see the greatest advances coming from A.I. and human cooperation. The potential is unlimited and artificial intelligence machines could be used in the near future to provide services in a number of sectors. What the future holds in this exciting field, remains to be seen!
Kumar, A., Sharma, A., &Goswami, A. (2014) Robotics & Artificial Intelligence. International Journal of Innovative Research in Modern World (IJIRMW), 1(2), pp.1-6.
Ingrand, F., &Ghallab, M. (2014) Robotics and artificial intelligence: a perspective on deliberation functions. AI Commun. 27(1), pp.63–80.